This content originally appeared on Level Up Coding - Medium and was authored by Mykhailo Kushnir
Scientifically proven insights. Sort of.
I'm an enthusiast of machine learning, and Medium has been a source of great content for me for years. During last year I've become a writer myself, and naturally, I wondered how to get better at it. As usual, I took a data-driven approach.
For this task, I've tried to explore headlines from relevant Machine Learning tag. I wondered who the most popular authors are and what topics they raise. In the end, I've discovered that there are distinctive patterns in headlines, and they probably influence the success of the post.
Disclaimers: I’ve used titles of posts (+ some metadata) posted in ML tag at Medium during 2021. They are publicly available to all Medium subscribers.
I'm using claps as a proxy for the post's success. While Medium pays on a view-based approach, claps are still relevant and represent recognition of the material. At least, that's true for my own works.
Most popular authors
Let's start with some recognition. If you are a vivid reader of ML posts, you'll probably notice some familiar names.
… And the first place goes to Natassha Selvaraj, as she achieved 485 claps per post on average. The second place raises some questions, though. While for most authors, the amount of claps are correlated with the number of responses and number of followers, I cannot observe the same pattern for this author. I want to point out that here and beyond, I'll not express criticism over the approach my fellow bloggers take to market their content. My goal is to share simple statistical facts. Let us start with those.
Lesson Learned: there seems to be a dependency between quality and quantity. Authors that posted more than 20 posts have 20 more claps in average, than those who posted less (71 clap vs 51 clap). If you want to write a good headline - make attempts!
Note: I've used only authors with 20 or more posts during last year in ML tag at Medium for this section. The goal was to skip outliers that reached higher levels of claps for one or two posts. I'm convinced that a good author consistently posts quality content, and my analysis was performed with those thoughts.
Most frequent authors
While I believe that quality content takes time and therefore it’s hard to provide lots of it during one year, I wanted to give recognition to those capable of posting lots of material as well. Besides, most of the blogs revealed above are not personal pages but rather a publication. For example, Technical Source, Synced, ODSC, TDS Editors and others.
Lesson Learned: Cooperation helps with the frequency of published content. Search for like-minded people who can go on the writers path along with you.
Most popular posts
Please note that Medium‘s using units for a number of claps beyond 1000, hence such “perfect” amount of claps you see in the table. Still, it gives us the ability to decide the best posts of 2021 in the Machine Learning subject.
Now take a distant glance at this list. Do you see any distinctive patterns? First of all, they are clearly pretty catchy. Most of the authors are either using the “How to” style or using numbers in the first part of the headline to underline value for a future reader.
Lesson Learned: some basic headline writing rules are still in the game and probably would be here for a time, as they are strongly tied to humans psychology.
Topics of most popular posts
This screenshot suggests that most of the authors are using relevant keywords in their headlines. One good hypothesis you can build around it is that they aim to leverage the power of SEO for the sake of traffic. Search engines value text in headlines more than text in the body of the post, so it’s quite reasonable. As you can see, I’ve not only learned that lesson but actually implemented it :)
Patterns in headlines
Having numbers at the start of the headline actually cooccurs with more claps for the post. At first, I thought it would be explained by the general quality of authors with high numbers of claps, but this trend holds even in those less successful
Lesson Learned: Don’t shy away from using “Here’s X points” or “How to” headlines style. While they might appear cheesy at first glance, they have certain functions and can be good for inexperienced authors. You’ll develop your own style with time, but basics are basics for reasons.
Summary
Machine Learning headlines abide by many common rules like SEO appliance, how-to structure, “catchiness”. In order to be successful, you need to comply with them but also add a drip of your own style. Most prominent authors either represent a great company they’re working in or have some distinctive experience in the world of ML. Don’t be shy even if you are not one of those two. Your unique learning experience or a nice pet project can be something are searching for!
How to write a good Machine Learning headline on Medium was originally published in Level Up Coding on Medium, where people are continuing the conversation by highlighting and responding to this story.
This content originally appeared on Level Up Coding - Medium and was authored by Mykhailo Kushnir
Mykhailo Kushnir | Sciencx (2022-02-05T01:16:35+00:00) How to write a good Machine Learning headline on Medium. Retrieved from https://www.scien.cx/2022/02/05/how-to-write-a-good-machine-learning-headline-on-medium/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.